Hi Carl,
> Manfred, my reply was assuming use at high power since this is not a
> QRP or receiver forum. In that case iron powder is the only accepted
> material as has been proven time and time.
A few years ago I needed to make a balun for 160 to 10 meters, that
should be able to work at 1.5kW on a transmission line that could have
high SWR. After comparing different powdered iron and ferrite materials,
I settled for 61-type ferrite for the core. It provided the most
cost-effective solution, to fulfill the requirements in terms of loss,
distortion, and bandwidth. I tried to squeeze #2 powdered iron into this
application, because it's significantly cheaper than #61 ferrite, but I
found no way to achieve the required performance with this material. So
I can't agree with iron powder being the only accepted material. At
least in this particular case, I found ferrite to be much superior.
> Even Sevik was forced to back down from his ferrite obsession when he
> was challenged by many and it became obvious that he never tested his
> designs at power.
The only thing I know about Sevik is his book about baluns, so I don't
know of his trends over time! Maybe I will come back to iron powder
someday too, but since having some meltdowns with iron powder, I'm a
firm fan of ferrite, for most applications including low and high power
broadband transformers! Which does not mean that I would think it
impossible to use powdered iron, but just in my experience ferrite
allows the more effective solution.
Jim,
> I've studied ferrites a LOT with respect to RFI suppression and common
> mode chokes, but I'm not an expert on their use in transformers. That said,
> there are some ferrite materials that offer VERY low losses at HF, most
> notably Fair-Rite #61 and #67.
Some ferrite materials have low losses even up into the UHF range. To my
knowledge, no powdered iron material achieves that.
> W8JI has observed that if the transformer is
> NOT handling power (for example, a Beverage transformer), a lossy material
> may be more suitable, depending on its mu over the frequency range of
> interest.
Lossy material might be more acceptable there, but never preferrable
over low loss material. But there is no need to bother with lossy
material, at least not at HF! We have it if we want it, such as the 73
material which is great for soaking up HF and thus cleaning up noisy
switching power supplies, but we also have low loss materials such as
the 61 and 67 you mention.
> Another VERY important point with respect to mu (permeability). When single
> numbers are quoted, they are for the mu at low frequencies. But the mu of ALL
> of these materials varies over a wide range as a function of frequency. Low
> loss materials like Fair-Rite #61, #67, and #68 have relatively low values of
> mu, but #61 maintains its value to about 20MHz, and the other two hold theirs
> to nearly 100 MHz.
I didn't know they had so much variation with frequency, but also the
permeability varies with temperature and with the magnetization (or flux
density) level. So, most certainly, ungapped ferrites have no place
where a constant, stable permeability is needed. In such applications,
either a ferrite core with air gap, or a powdered iron core is needed.
> By contrast, high loss materials designed for suppression,
> like #43 and #31 start out with low frequency mu values of 800 and 1200
> respectively, but have dropped to 200 at 10 MHz and 100 at 100 MHz. More
> important, #43 is more lossy than inductive above 10 MHz; for #31, that
> transition is around 4 MHz.
Yes, and for #73, it's even lower. In fact, many designers of switching
power supplies liberally install #43 material beads in their products,
in an attempt to absorb noise. It's a pretty futile attempt for us hams,
because #43 will be effective only in the highest HF range, and at VHF,
in absorbing noise! The simple step of using #73 material for noise
absorption made my 40 Ampere switcher so RF-quiet that it has since been
quoted as an example in several books and magazines!
Of course, don't ever wind a broadband transformer for HF on #73
material! It would make a good heater and not much else! :-)
Carl,
> In a broadband low level receiver or transmitter path core losses can easily
> be recovered with more gain.
Only as long as the signal hasn't dropped below the noise floor of the
following gain stage!
> However the smaller you get with the ferrite
> the more prone they become to non linearity events, primarily in receivers.
That's indeed a big problem, and with transmitters it's a problem too.
Perhaps not for the user, but sure enough for his neighbors! :-)
Both iron powder and ferrite cores have to be driven at pretty low
levels to keep nonlinearity effects in the green range. But then, at RF
anyway they have to be driven pretty low to keep them from overheating!
At the flux density levels you can safely put into a core in the upper
HF range, be it iron powder or ferrite, without exceeding a manageable
loss, the distortion from nonlinearity will be acceptable in almost all
cases.
And now a question for those among you with good ideas: Can anyone come
up with a way to make a broadband, high impedance transformer? Say,
something to convert from 3 kiloohm to 50 Ohm, covering 160 to 10
meters? The application is obvious, a broadband no-tune tube type power
amplifier. My problem is, I can't find a way to make such a transformer,
because wire length is excessive compared to the wavelength, and
interwinding capacitance is a serious problem. I assume that nobody has
been able to do this, and that this is the reason for the lack of
factory-made no-tune HF power amps using tubes. Anyone who manages to
build a suitable transformer would become famous, at least among us HF
power freaks! :-)
Manfred.
========================
Visit my hobby homepage!
http://ludens.cl
========================
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|